Cell factory design with advanced metabolic modelling empowered by artificial intelligence.

Advances in synthetic biology and artificial intelligence (AI) have provided new opportunities for modern biotechnology. High-performance cell factories, the backbone of industrial biotechnology, are ultimately responsible for determining whether a bio-based product succeeds or fails in the fierce competition with petroleum-based products. To date, one of the greatest challenges in synthetic biology is the creation of high-performance cell factories in a consistent and efficient manner. As so-called white-box models, numerous metabolic network models have been developed and used in computational strain design. Moreover, great progress has been made in AI-powered strain engineering in recent years. Both approaches have advantages and disadvantages. Therefore, the deep integration of AI with metabolic models is crucial for the construction of superior cell factories with higher titres, yields and production rates. The detailed applications of the latest advanced metabolic models and AI in computational strain design are summarized in this review. Additionally, approaches for the deep integration of AI and metabolic models are discussed. It is anticipated that advanced mechanistic metabolic models powered by AI will pave the way for the efficient construction of powerful industrial chassis strains in the coming years.

An NADH/NAD+-favored aldo-keto reductase facilitates avilamycin A biosynthesis by primarily catalyzing oxidation of avilamycin C.

Avilamycins, which possess potent inhibitory activity against Gram-positive bacteria, are a group of oligosaccharide antibiotics produced by Streptomyces viridochromogenes. Among these structurally related oligosaccharide antibiotics, avilamycin A serves as the main bioactive component in veterinary drugs and animal feed additives, which differs from avilamycin C only in the redox state of the two-carbon branched-chain of the terminal octose moiety. However, the mechanisms underlying assembly and modification of the oligosaccharide chain to diversify individual avilamycins remain poorly understood. Here, we report that AviZ1, an aldo-keto reductase in the avilamycin pathway, can catalyze the redox conversion between avilamycins A and C. Remarkably, the ratio of these two components produced by AviZ1 depends on the utilization of specific redox cofactors, namely NADH/NAD+ or NADPH/NADP+. These findings are inspired by gene disruption and complementation experiments and are further supported by in vitro enzymatic activity assays, kinetic analyses, and cofactor affinity studies on AviZ1-catalyzed redox reactions. Additionally, the results from sequence analysis, structure prediction, and site-directed mutagenesis of AviZ1 validate it as an NADH/NAD+-favored aldo-keto reductase that primarily oxidizes avilamycin C to form avilamycin A by utilizing abundant NAD+ in vivo. Building upon the biological function and catalytic activity of AviZ1, overexpressing AviZ1 in S. viridochromogenes is thus effective to improve the yield and proportion of avilamycin A in the fermentation profile of avilamycins. This study represents, to our knowledge, the first characterization of biochemical reactions involved in avilamycin biosynthesis and contributes to the construction of high-performance strains with industrial value.IMPORTANCEAvilamycins are a group of oligosaccharide antibiotics produced by Streptomyces viridochromogenes, which can be used as veterinary drugs and animal feed additives. Avilamycin A is the most bioactive component, differing from avilamycin C only in the redox state of the two-carbon branched-chain of the terminal octose moiety. Currently, the biosynthetic pathway of avilamycins is not clear. Here, we report that AviZ1, an aldo-keto reductase in the avilamycin pathway, can catalyze the redox conversion between avilamycins A and C. More importantly, AviZ1 exhibits a unique NADH/NAD+ preference, allowing it to efficiently catalyze the oxidation of avilamycin C to form avilamycin A using abundant NAD+ in cells. Thus, overexpressing AviZ1 in S. viridochromogenes is effective to improve the yield and proportion of avilamycin A in the fermentation profile of avilamycins. This study serves as an enzymological guide for rational strain design, and the resulting high-performance strains have significant industrial value.

Itaconic acid production by co-feeding of Ustilago maydis: A combined approach of experimental data, design of experiments, and metabolic modeling.

Itaconic acid is a platform chemical with a range of applications in polymer synthesis and is also discussed for biofuel production. While produced in industry from glucose or sucrose, co-feeding of glucose and acetate was recently discussed to increase itaconic acid production by the smut fungus Ustilago maydis. In this study, we investigate the optimal co-feeding conditions by interlocking experimental and computational methods. Flux balance analysis indicates that acetate improves the itaconic acid yield up to a share of 40% acetate on a carbon molar basis. A design of experiment results in the maximum yield of 0.14 itaconic acid per carbon source from 100 g L - 1 $\,\text{g L}{}^{-1}$ glucose and 12 g L - 1 $\,\text{g L}{}^{-1}$ acetate. The yield is improved by around 22% when compared to feeding of glucose as sole carbon source. To further improve the yield, gene deletion targets are discussed that were identified using the metabolic optimization tool OptKnock. The study contributes ideas to reduce land use for biotechnology by incorporating acetate as co-substrate, a C2-carbon source that is potentially derived from carbon dioxide.

Introduction of NADH-dependent nitrate assimilation in Synechococcus sp. PCC 7002 improves photosynthetic production of 2-methyl-1-butanol and isobutanol.

Cyanobacteria hold promise for renewable chemical production due to their photosynthetic nature, but engineered strains frequently display poor production characteristics. These difficulties likely arise in part due to the distinctive photoautotrophic metabolism of cyanobacteria. In this work, we apply a genome-scale metabolic model of the cyanobacteria Synechococus sp. PCC 7002 to identify strain designs accounting for this unique metabolism that are predicted to improve the production of various biofuel alcohols (e.g. 2-methyl-1-butanol, isobutanol, and 1-butanol) synthesized via an engineered biosynthesis pathway. Using the model, we identify that the introduction of a large, non-native NADH-demand into PCC 7002's metabolic network is predicted to enhance production of these alcohols by promoting NADH-generating reactions upstream of the production pathways. To test this, we construct strains of PCC 7002 that utilize a heterologous, NADH-dependent nitrite reductase in place of the native, ferredoxin-dependent enzyme to create an NADH-demand in the cells when grown on nitrate-containing media. We find that photosynthetic production of both isobutanol and 2-methyl-1-butanol is significantly improved in the engineered strain background relative to that in a wild-type background. We additionally identify that the use of high-nutrient media leads to a substantial prolongment of the production curve in our alcohol production strains. The metabolic engineering strategy identified and tested in this work presents a novel approach to engineer cyanobacterial production strains that takes advantage of a unique aspect of their metabolism and serves as a basis on which to further develop strains with improved production of these alcohols and related products.

Integrated knowledge mining, genome-scale modeling, and machine learning for predicting Yarrowia lipolytica bioproduction.

Predicting bioproduction titers from microbial hosts has been challenging due to complex interactions between microbial regulatory networks, stress responses, and suboptimal cultivation conditions. This study integrated knowledge mining, feature extraction, genome-scale modeling (GSM), and machine learning (ML) to develop a model for predicting Yarrowia lipolytica chemical titers (i.e., organic acids, terpenoids, etc.). First, Y. lipolytica production data, including cultivation conditions, genetic engineering strategies, and product information, was manually collected from literature (~100 papers) and stored as either numerical (e.g., substrate concentrations) or categorical (e.g., bioreactor modes) variables. For each case recorded, central pathway fluxes were estimated using GSMs and flux balance analysis (FBA) to provide metabolic features. Second, a ML ensemble learner was trained to predict strain production titers. Accurate predictions on the test data were obtained for instances with production titers >1 g/L (R2 = 0.87). However, the model had reduced predictability for low performance strains (0.01-1 g/L, R2 = 0.29) potentially due to biosynthesis bottlenecks not captured in the features. Feature ranking indicated that the FBA fluxes, the number of enzyme steps, the substrate inputs, and thermodynamic barriers (i.e., Gibbs free energy of reaction) were the most influential factors. Third, the model was evaluated on other oleaginous yeasts and indicated there were conserved features for some hosts that can be potentially exploited by transfer learning. The platform was also designed to assist computational strain design tools (such as OptKnock) to screen genetic targets for improved microbial production in light of experimental conditions.

Genome-scale metabolic modeling of P. thermoglucosidasius NCIMB 11955 reveals metabolic bottlenecks in anaerobic metabolism.

Parageobacillus thermoglucosidasius represents a thermophilic, facultative anaerobic bacterial chassis, with several desirable traits for metabolic engineering and industrial production. To further optimize strain productivity, a systems level understanding of its metabolism is needed, which can be facilitated by a genome-scale metabolic model. Here, we present p-thermo, the most complete, curated and validated genome-scale model (to date) of Parageobacillus thermoglucosidasius NCIMB 11955. It spans a total of 890 metabolites, 1175 reactions and 917 metabolic genes, forming an extensive knowledge base for P. thermoglucosidasius NCIMB 11955 metabolism. The model accurately predicts aerobic utilization of 22 carbon sources, and the predictive quality of internal fluxes was validated with previously published 13C-fluxomics data. In an application case, p-thermo was used to facilitate more in-depth analysis of reported metabolic engineering efforts, giving additional insight into fermentative metabolism. Finally, p-thermo was used to resolve a previously uncharacterised bottleneck in anaerobic metabolism, by identifying the minimal required supplemented nutrients (thiamin, biotin and iron(III)) needed to sustain anaerobic growth. This highlights the usefulness of p-thermo for guiding the generation of experimental hypotheses and for facilitating data-driven metabolic engineering, expanding the use of P. thermoglucosidasius as a high yield production platform.

Speeding up the core algorithm for the dual calculation of minimal cut sets in large metabolic networks.

BACKGROUND: The concept of minimal cut sets (MCS) has become an important mathematical framework for analyzing and (re)designing metabolic networks. However, the calculation of MCS in genome-scale metabolic models is a complex computational problem. The development of duality-based algorithms in the last years allowed the enumeration of thousands of MCS in genome-scale networks by solving mixed-integer linear problems (MILP). A recent advancement in this field was the introduction of the MCS2 approach. In contrast to the Farkas-lemma-based dual system used in earlier studies, the MCS2 approach employs a more condensed representation of the dual system based on the nullspace of the stoichiometric matrix, which, due to its reduced dimension, holds promise to further enhance MCS computations. RESULTS: In this work, we introduce several new variants and modifications of duality-based MCS algorithms and benchmark their effects on the overall performance. As one major result, we generalize the original MCS2 approach (which was limited to blocking the operation of certain target reactions) to the most general case of MCS computations with arbitrary target and desired regions. Building upon these developments, we introduce a new MILP variant which allows maximal flexibility in the formulation of MCS problems and fully leverages the reduced size of the nullspace-based dual system. With a comprehensive set of benchmarks, we show that the MILP with the nullspace-based dual system outperforms the MILP with the Farkas-lemma-based dual system speeding up MCS computation with an averaged factor of approximately 2.5. We furthermore present several simplifications in the formulation of constraints, mainly related to binary variables, which further enhance the performance of MCS-related MILP. However, the benchmarks also reveal that some highly condensed formulations of constraints, especially on reversible reactions, may lead to worse behavior when compared to variants with a larger number of (more explicit) constraints and involved variables. CONCLUSIONS: Our results further enhance the algorithmic toolbox for MCS calculations and are of general importance for theoretical developments as well as for practical applications of the MCS framework.

Optimization of Yarrowia lipolytica-based consolidated biocatalyst through synthetic biology approach: transcription units and signal peptides shuffling.

Nowadays considerable effort is being pursued towards development of consolidated microbial biocatalysts that will be able to utilize complex, non-pretreated substrates and produce valuable compounds. In such engineered microbes, synthesis of extracellular hydrolases may be fine-tuned by different approaches, like strength of promoter, type of secretory tag, and gene copy number. In this study, we investigated if organization of a multi-element expression cassette impacts the resultant Yarrowia lipolytica transformants' phenotype, presuming that different variants of the cassette are composed of the same regulatory elements and encode the same mature proteins. To this end, Y. lipolytica cells were transformed with expression cassettes bearing a pair of genes encoding exactly the same mature amylases, but fused to four different signal peptides (SP), and located interchangeably in either first or second position of a synthetic DNA construction. The resultant strains were tested for growth on raw and pretreated complex substrates of different plant origin for comprehensive examination of the strains' acquired characteristics. Optimized strain was tested in batch bioreactor cultivations for growth and lipids accumulation. Based on the conducted research, we concluded that the positional order of transcription units (TU) and the type of exploited SP affect final characteristics of the resultant consolidated biocatalyst strains, and thus could be considered as additional factors to be evaluated upon consolidated biocatalysts optimization. KEY POINTS: • Y. lipolytica growing on raw starch was constructed and tested on different substrates. • Impact of expression cassette design and SP on biocatalysts' phenotype was evidenced. • Consolidated biocatalyst process for lipids production from starch was conducted.

Comparison of pathway analysis and constraint-based methods for cell factory design.

BACKGROUND: Computational strain optimisation methods (CSOMs) have been successfully used to exploit genome-scale metabolic models, yielding strategies useful for allowing compound overproduction in metabolic cell factories. Minimal cut sets are particularly interesting since their definition allows searching for intervention strategies that impose strong growth-coupling phenotypes, and are not subject to optimality bias when compared with simulation-based CSOMs. However, since both types of methods have different underlying principles, they also imply different ways to formulate metabolic engineering problems, posing an obstacle when comparing their outputs. RESULTS: In this work, we perform an in-depth analysis of potential strategies that can be obtained with both methods, providing a critical comparison of performance, robustness, predicted phenotypes as well as strategy structure and size. To this end, we devised a pipeline including enumeration of strategies from evolutionary algorithms (EA) and minimal cut sets (MCS), filtering and flux analysis of predicted mutants to optimize the production of succinic acid in Saccharomyces cerevisiae. We additionally attempt to generalize problem formulations for MCS enumeration within the context of growth-coupled product synthesis. Strategies from evolutionary algorithms show the best compromise between acceptable growth rates and compound overproduction. However, constrained MCSs lead to a larger variety of phenotypes with several degrees of growth-coupling with production flux. The latter have proven useful in revealing the importance, in silico, of the gamma-aminobutyric acid shunt and manipulation of cofactor pools in growth-coupled designs for succinate production, mechanisms which have also been touted as potentially useful for metabolic engineering. CONCLUSIONS: The two main groups of CSOMs are valuable for finding growth-coupled mutants. Despite the limitations in maximum growth rates and large strategy sizes, MCSs help uncover novel mechanisms for compound overproduction and thus, analyzing outputs from both methods provides a richer overview on strategies that can be potentially carried over in vivo.

Yeast Genome-Scale Metabolic Models for Simulating Genotype-Phenotype Relations.

Understanding genotype-phenotype dependency is a universal aim for all life sciences. While the complete genotype-phenotype relations remain challenging to resolve, metabolic phenotypes are moving within the reach through genome-scale metabolic model simulations. Genome-scale metabolic models are available for commonly investigated yeasts, such as model eukaryote and domesticated fermentation species Saccharomyces cerevisiae, and automatic reconstruction methods facilitate obtaining models for any sequenced species. The models allow for investigating genotype-phenotype relations through simulations simultaneously considering the effects of nutrient availability, and redox and energy homeostasis in cells. Genome-scale models also offer frameworks for omics data integration to help to uncover how the translation of genotypes to the apparent phenotypes is regulated at different levels. In this chapter, we provide an overview of the yeast genome-scale metabolic models and the simulation approaches for using these models to interrogate genotype-phenotype relations. We review the methodological approaches according to the underlying biological reasoning in order to inspire formulating novel questions and applications that the genome-scale metabolic models could contribute to. Finally, we discuss current challenges and opportunities in the genome-scale metabolic model simulations.

In silico strategies to couple production of bioethanol with growth in cyanobacteria.

Cyanobacteria have been considered as promising candidates for sustainable bioproduction from inexpensive raw materials, as they grow on light, carbon dioxide, and minimal inorganic nutrients. In this study, we present a genome-scale metabolic network model for Synechocystis sp. PCC 6803 and study the optimal design of the strain for ethanol production by using a mixed integer linear problem reformulation of a bilevel programming problem that identifies gene knockouts which lead to coupling between growth and product synthesis. Five mutants were found, where the in silico model predicts coupling between biomass growth and ethanol production in photoautotrophic conditions. The best mutant gives an in silico ethanol production of 1.054 mmol·gDW -1 ·h -1 .

A mathematical framework for yield (vs. rate) optimization in constraint-based modeling and applications in metabolic engineering.

BACKGROUND: The optimization of metabolic rates (as linear objective functions) represents the methodical core of flux-balance analysis techniques which have become a standard tool for the study of genome-scale metabolic models. Besides (growth and synthesis) rates, metabolic yields are key parameters for the characterization of biochemical transformation processes, especially in the context of biotechnological applications. However, yields are ratios of rates, and hence the optimization of yields (as nonlinear objective functions) under arbitrary linear constraints is not possible with current flux-balance analysis techniques. Despite the fundamental importance of yields in constraint-based modeling, a comprehensive mathematical framework for yield optimization is still missing. RESULTS: We present a mathematical theory that allows one to systematically compute and analyze yield-optimal solutions of metabolic models under arbitrary linear constraints. In particular, we formulate yield optimization as a linear-fractional program. For practical computations, we transform the linear-fractional yield optimization problem to a (higher-dimensional) linear problem. Its solutions determine the solutions of the original problem and can be used to predict yield-optimal flux distributions in genome-scale metabolic models. For the theoretical analysis, we consider the linear-fractional problem directly. Most importantly, we show that the yield-optimal solution set (like the rate-optimal solution set) is determined by (yield-optimal) elementary flux vectors of the underlying metabolic model. However, yield- and rate-optimal solutions may differ from each other, and hence optimal (biomass or product) yields are not necessarily obtained at solutions with optimal (growth or synthesis) rates. Moreover, we discuss phase planes/production envelopes and yield spaces, in particular, we prove that yield spaces are convex and provide algorithms for their computation. We illustrate our findings by a small example and demonstrate their relevance for metabolic engineering with realistic models of E. coli. CONCLUSIONS: We develop a comprehensive mathematical framework for yield optimization in metabolic models. Our theory is particularly useful for the study and rational modification of cell factories designed under given yield and/or rate requirements.

In silico model-guided identification of transcriptional regulator targets for efficient strain design.

BACKGROUND: Cellular metabolism is tightly regulated by hard-wired multiple layers of biological processes to achieve robust and homeostatic states given the limited resources. As a result, even the most intuitive enzyme-centric metabolic engineering endeavours through the up-/down-regulation of multiple genes in biochemical pathways often deliver insignificant improvements in the product yield. In this regard, targeted engineering of transcriptional regulators (TRs) that control several metabolic functions in modular patterns is an interesting strategy. However, only a handful of in silico model-added techniques are available for identifying the TR manipulation candidates, thus limiting its strain design application. RESULTS: We developed hierarchical-Beneficial Regulatory Targeting (h-BeReTa) which employs a genome-scale metabolic model and transcriptional regulatory network (TRN) to identify the relevant TR targets suitable for strain improvement. We then applied this method to industrially relevant metabolites and cell factory hosts, Escherichia coli and Corynebacterium glutamicum. h-BeReTa suggested several promising TR targets, many of which have been validated through literature evidences. h-BeReTa considers the hierarchy of TRs in the TRN and also accounts for alternative metabolic pathways which may divert flux away from the product while identifying suitable metabolic fluxes, thereby performing superior in terms of global TR target identification. CONCLUSIONS: In silico model-guided strain design framework, h-BeReTa, was presented for identifying transcriptional regulator targets. Its efficacy and applicability to microbial cell factories were successfully demonstrated via case studies involving two cell factory hosts, as such suggesting several intuitive targets for overproducing various value-added compounds.

Multilevel engineering of the upstream module of aromatic amino acid biosynthesis in Saccharomyces cerevisiae for high production of polymer and drug precursors.

A multilevel approach was implemented in Saccharomyces cerevisiae to optimize the precursor module of the aromatic amino acid biosynthesis pathway, which is a rich resource for synthesizing a great variety of chemicals ranging from polymer precursor, to nutraceuticals and pain-relief drugs. To facilitate the discovery of novel targets to enhance the pathway flux, we incorporated the computational tool YEASTRACT for predicting novel transcriptional repressors and OptForce strain-design for identifying non-intuitive pathway interventions. The multilevel approach consisted of (i) relieving the pathway from strong transcriptional repression, (ii) removing competing pathways to ensure high carbon capture, and (iii) rewiring precursor pathways to increase the carbon funneling to the desired target. The combination of these interventions led to the establishment of a S. cerevisiae strain with shikimic acid (SA) titer reaching as high as 2.5gL-1, 7-fold higher than the base strain. Further expansion of the platform led to the titer of 2.7gL-1 of muconic acid (MA) and its intermediate protocatechuic acid (PCA) together. Both the SA and MA production platforms demonstrated increases in titer and yield nearly 300% from the previously reported, highest-producing S. cerevisiae strains. Further examination elucidated the diverged impacts of disrupting the oxidative branch (ZWF1) of the pentose phosphate pathway on the titers of desired products belonging to different portions of the pathway. The investigation of other non-intuitive interventions like the deletion of the Pho13 enzyme also revealed the important role of the transaldolase in determining the fate of the carbon flux in the pathways of study. This integrative approach identified novel determinants at both transcriptional and metabolic levels that constrain the flux entering the aromatic amino acid pathway. In the future, this platform can be readily used for engineering the downstream modules toward the production of important plant-sourced aromatic secondary metabolites.

Flux control-based design of furfural-resistance strains of Saccharomyces cerevisiae for lignocellulosic biorefinery.

We have previously developed a dynamic flux balance analysis of Saccharomyces cerevisiae for elucidation of genome-wide flux response to furfural perturbation (Unrean and Franzen, Biotechnol J 10(8):1248-1258, 2015). Herein, the dynamic flux distributions were analyzed by flux control analysis to identify target overexpressed genes for improved yeast robustness against furfural. The flux control coefficient (FCC) identified overexpressing isocitrate dehydrogenase (IDH1), a rate-controlling flux for ethanol fermentation, and dicarboxylate carrier (DIC1), a limiting flux for cell growth, as keys of furfural-resistance phenotype. Consistent with the model prediction, strain characterization showed 1.2- and 2.0-fold improvement in ethanol synthesis and furfural detoxification rates, respectively, by IDH1 overexpressed mutant compared to the control. DIC1 overexpressed mutant grew at 1.3-fold faster and reduced furfural at 1.4-fold faster than the control under the furfural challenge. This study hence demonstrated the FCC-based approach as an effective tool for guiding the design of robust yeast strains.

A method for analysis and design of metabolism using metabolomics data and kinetic models: Application on lipidomics using a novel kinetic model of sphingolipid metabolism.

We present a model-based method, designated Inverse Metabolic Control Analysis (IMCA), which can be used in conjunction with classical Metabolic Control Analysis for the analysis and design of cellular metabolism. We demonstrate the capabilities of the method by first developing a comprehensively curated kinetic model of sphingolipid biosynthesis in the yeast Saccharomyces cerevisiae. Next we apply IMCA using the model and integrating lipidomics data. The combinatorial complexity of the synthesis of sphingolipid molecules, along with the operational complexity of the participating enzymes of the pathway, presents an excellent case study for testing the capabilities of the IMCA. The exceptional agreement of the predictions of the method with genome-wide data highlights the importance and value of a comprehensive and consistent engineering approach for the development of such methods and models. Based on the analysis, we identified the class of enzymes regulating the distribution of sphingolipids among species and hydroxylation states, with the D-phospholipase SPO14 being one of the most prominent. The method and the applications presented here can be used for a broader, model-based inverse metabolic engineering approach.

Model-based metabolic engineering enables high yield itaconic acid production by Escherichia coli.

Itaconic acid is a high potential platform chemical which is currently industrially produced by Aspergillus terreus. Heterologous production of itaconic acid with Escherichia coli could help to overcome limitations of A. terreus regarding slow growth and high sensitivity to oxygen supply. However, the performance achieved so far with E. coli strains is still low. We introduced a plasmid (pCadCS) carrying genes for itaconic acid production into E. coli and applied a model-based approach to construct a high yield production strain. Based on the concept of minimal cut sets, we identified intervention strategies that guarantee high itaconic acid yield while still allowing growth. One cut set was selected and the corresponding genes were iteratively knocked-out. As a conceptual novelty, we pursued an adaptive approach allowing changes in the model and initially calculated intervention strategy if a genetic modification induces changes in byproduct formation. Using this approach, we iteratively implemented five interventions leading to high yield itaconic acid production in minimal medium with glucose as substrate supplemented with small amounts of glutamic acid. The derived E. coli strain (ita23: MG1655 ∆aceA ∆sucCD ∆pykA ∆pykF ∆pta ∆Picd::cam_BBa_J23115 pCadCS) synthesized 2.27g/l itaconic acid with an excellent yield of 0.77mol/(mol glucose). In a fed-batch cultivation, this strain produced 32g/l itaconic acid with an overall yield of 0.68mol/(mol glucose) and a peak productivity of 0.45g/l/h. These values are by far the highest that have ever been achieved for heterologous itaconic acid production and indicate that realistic applications come into reach.

Transcriptomics-based strain optimization tool for designing secondary metabolite overproducing strains of Streptomyces coelicolor.

In silico model-driven analysis using genome-scale model of metabolism (GEM) has been recognized as a promising method for microbial strain improvement. However, most of the current GEM-based strain design algorithms based on flux balance analysis (FBA) heavily rely on the steady-state and optimality assumptions without considering any regulatory information. Thus, their practical usage is quite limited, especially in its application to secondary metabolites overproduction. In this study, we developed a transcriptomics-based strain optimization tool (tSOT) in order to overcome such limitations by integrating transcriptomic data into GEM. Initially, we evaluated existing algorithms for integrating transcriptomic data into GEM using Streptomyces coelicolor dataset, and identified iMAT algorithm as the only and the best algorithm for characterizing the secondary metabolism of S. coelicolor. Subsequently, we developed tSOT platform where iMAT is adopted to predict the reaction states, and successfully demonstrated its applicability to secondary metabolites overproduction by designing actinorhodin (ACT), a polyketide antibiotic, overproducing strain of S. coelicolor. Mutants overexpressing tSOT targets such as ribulose 5-phosphate 3-epimerase and NADP-dependent malic enzyme showed 2 and 1.8-fold increase in ACT production, thereby validating the tSOT prediction. It is expected that tSOT can be used for solving other metabolic engineering problems which could not be addressed by current strain design algorithms, especially for the secondary metabolite overproductions.

Quantitative prediction of genome-wide resource allocation in bacteria.

Predicting resource allocation between cell processes is the primary step towards decoding the evolutionary constraints governing bacterial growth under various conditions. Quantitative prediction at genome-scale remains a computational challenge as current methods are limited by the tractability of the problem or by simplifying hypotheses. Here, we show that the constraint-based modeling method Resource Balance Analysis (RBA), calibrated using genome-wide absolute protein quantification data, accurately predicts resource allocation in the model bacterium Bacillus subtilis for a wide range of growth conditions. The regulation of most cellular processes is consistent with the objective of growth rate maximization except for a few suboptimal processes which likely integrate more complex objectives such as coping with stressful conditions and survival. As a proof of principle by using simulations, we illustrated how calibrated RBA could aid rational design of strains for maximizing protein production, offering new opportunities to investigate design principles in prokaryotes and to exploit them for biotechnological applications.

Characterizing metabolic pathway diversification in the context of perturbation size.

Cell metabolism is an important platform for sustainable biofuel, chemical and pharmaceutical production but its complexity presents a major challenge for scientists and engineers. Although in silico strains have been designed in the past with predicted performances near the theoretical maximum, real-world performance is often sub-optimal. Here, we simulate how strain performance is impacted when subjected to many randomly varying perturbations, including discrepancies between gene expression and in vivo flux, osmotic stress, and substrate uptake perturbations due to concentration gradients in bioreactors. This computational study asks whether robust performance can be achieved by adopting robustness-enhancing mechanisms from naturally evolved organisms-in particular, redundancy. Our study shows that redundancy, typically perceived as a ubiquitous robustness-enhancing strategy in nature, can either improve or undermine robustness depending on the magnitude of the perturbations. We also show that the optimal number of redundant pathways used can be predicted for a given perturbation size.